The value-ladenness of ancestry

Clustering humans based on their genetic ancestry is a common practice in human genomics. Genetically similar populations can be seen as statistical constructs that are labeled by population descriptors such as "race," "ethnicity," and "genetic ancestry." Recently, there has been a shift towards replacing the descriptor "race" with "genetic ancestry" because the latter is considered more objective. A descriptor is deemed objective if it adequately captures an underlying feature of the biological world, such as genetic similarities or differences between human sub-populations. However, claims of objectivity do not sufficiently explain the rationale for the choice and use of population descriptors such as "ancestry." This paper proposes an axiological approach to capture the choice and use of population descriptors in human genomics, by showing that the population descriptor "ancestry" is value-laden and that there is a legitimate role for values in the choice and use of population descriptors in genomics.